Life threatening myocardial ischaemia associated with the use of prostaglandin E1 to induce abortion


  • Uwe Schulte-Sasse

    Professor, Corresponding author
    1. Department of Anaesthesia and Critical Care Medicine, Community Hospital of the University Heidelberg, Heilbronn, Germany
      Correspondence: Professor U. Schulte-Sasse, Klinik für Anaesthesie und Operative Intensivmedizin, Klinikum Heilbronn GmbH, Am Gesundbrunnen, D-74064 Heilbronn, Germany.
    Search for more papers by this author

Correspondence: Professor U. Schulte-Sasse, Klinik für Anaesthesie und Operative Intensivmedizin, Klinikum Heilbronn GmbH, Am Gesundbrunnen, D-74064 Heilbronn, Germany.

Case reports

Case 1

At 10 weeks of gestation a 29 year old woman was advised to undergo therapeutic termination of pregnancy because of a long standing history of renal insufficiency. She had been treated for hypertension with β-adrenergic blocking agents and angiotensin converting enzyme inhibitors. Other risk factors included obesity (height 153 cm, weight 80 kg), hypercholesterolaemia (cholesterol between 239 and 300 mg/dL), and a history of smoking 20 cigarettes per day for years. Ten years previously she had suffered tightness of the chest and substernal pain in the presence of ‘known arrhythmia’ when 13 weeks pregnant. Five hours before the surgical procedure, a 1 mg gemeprost suppository (Cergem, Nourypharma, Oberschleiβheim, Germany) was inserted vaginally. Blood pressure, heart rate, electrocardiogram were not recorded during the time of uterine contractions and completion of abortion. On administration of anaesthesia, using thiopentone and 1% isoflurane, the electrocardiogram showed sinus tachycardia of 105 beats per minute with a blood pressure of 150/90 mmHg. After administration of the muscle relaxant vecuronium, her trachea was intubated and 50% oxygen and 50% nitrous oxide were used for ventilation. The oxygen saturation remained above 98%. At the end of the dilatation and evacuation, with a blood pressure of 110/60 mmHg and heart rate of 90 beats per minute a combination of 5 units of oxytocin and 0.5 mg methylergometrin (Syntometrin, Sandoz, Nürnberg, Germany) was injected intravenously, after which a laparoscopic tubal ligation was performed. At the end of the uneventful procedure, after discontinuing the inhalational anaesthetic with the patient breathing 100% oxygen, her blood pressure dropped to 90/50 mmHg and then became unrecordable. The heart rate fell to 40 beats per minute, and rapidly progressed to ectopic beats and ventricular fibrillation. A presumptive diagnosis of massive pulmonary embolism was made and systemic thrombolytic therapy using 1.5 × 106 units of streptokinase, accompanied by intravenous heparin (5000 units) was begun, after her blood pressure returned to perceptible levels. Over the next 30 minutes resuscitation with direct current shocks, atropine, and adrenaline resulted in a pulse rate of 120 beats per minute and a blood pressure of 130/60 mmHg. Blood pressure was maintained with adrenaline 7.5 μg/minute and norepinephrine 5 μg/minute. Emergency echocardiogram did not reveal a pulmonary embolus but indicated myocardial hypokinesis, most likely due to myocardial infarction. The patient was rushed to the cardiology laboratory, where complete occlusion of the ramus interventricularis anterior and an extended 80% stenosis of the right coronary artery was diagnosed by angiography. An emergency percutaneous transluminal coronary angioplasty re-established flow in the ramus interventricularis anterior, resulting in an immediate improvement of left ventricular contractility and the intravenous adrenergic support was stopped. After reopening the ramus interventricularis anterior, a 50% stenosis of the ramus diagnonalis could be demonstrated. One month later, after an uneventful recovery, apart from diffuse atherosclerotic plaques, no abnormal vessels or haemodynamically significant narrowing could be demonstrated by coronary artery angiography, supporting the suspicion that myocardial infarction was due to coronary spasm induced by administration of prostaglandin E1 (gemeprost).

Case 2

A 32 year old nulliparous woman presented for removal of the products of conception after fetal death had been diagnosed by pelvic ultrasound at 18 weeks of gestation. She was admitted for medical abortion. There was no history of pre-existing heart or central nervous system disease. She smoked eight cigarettes per day. The following morning a 1 mg gemeprost vaginal suppository was administered at 5:45 hours, and a second one at 13:00 hours. Blood pressure, heart rate, electrocardiogram were not recorded during the period of uterine contractions. At 14:17 hours the woman was found unconscious without respiration and dilated pupils. She was deeply cyanotic, with no measurable blood pressure. Peripheral pulses could not be palpated. Measures to resuscitate her were started, including intubation of the trachea and ventilation with 100% oxygen, intravenous administration of 1 mg adrenaline, crystalloid infusion, and low doses of dobutamine. These measures restored the heart beat, and she developed a systolic blood pressure of 80 mmHg gradually increasing over 20 minutes to 100 mmHg. The electrocardiogram showed sinus tachycardia of 110 beats per minute and ST depression on the precordial leads V3–6, indicating myocardial ischaemia. A presumptive diagnosis of acute coronary ischaemia due to prostaglandin-induced coronary arterial spasm was made. Heparin (5000 units) and 500 mg acetylsalycylic acid were administered intravenously, accompanied by low dose infusion of nitroglycerin. Serial testing revealed a maximum elevation of creatine kinase of 672 U/L, with MB isoenzyme increased to 66 U/L. Echocardio-graphy revealed impaired left ventricular function with an ejection fraction of 40% in the presence of 2.5 μg/kg/min dobutamine. At 15:55 hours, after expulsion of a macerated fetus, curettage was effected for uterine evacuation. She was then transferred to the cardiology laboratory. No overt signs of coronary artery disease could be detected, but the coronary angiogram of the left coronary artery demonstrated spasticity, mainly involving the circumflex artery. Seven hours later, when she was awake and haemodynamically stable, a diagnosis of right sided hemiparesis and motor aphasia was made. By the following day an incomplete homonymous hemianopia was recognised and the diagnosis of an ‘infarction of the left middle cerebral artery’ was made clinically and confirmed by magnetic resonance imaging. By noninvasive tests (carotid and transcranial Doppler ultrasonography and magnetic resonance angiography) 24 hours after cardiovascular collapse no signs of pre-existing cerebrovascular disease could be detected. After 20 days in hospital, she was transferred to a rehabilitation institution. After two months she was discharged, still with rehabilitation help. At the end of her stay in the hospital her ability to walk, to use her right arm and hand and to speak had improved significantly, but she is unable to work and remains severely handicapped.


Prostaglandin E1 (gemeprost) is an abortifacient that, unlike prostaglandin E21–4 has not been reported to cause coronary artery spasm or myocardial infarction. A literature search that included Medline (1976 to present) identified only two reports of transient cardiovascular morbidity following gemeprost, but—in contrast to prostaglandin E2–neither described myocardial ischaemia5,6. An unpublished database of gemeprost's manufacturer (ONO Pharmaceutical Co Ltd) contains three cases in which coronary artery spasm could have been the cause of serious morbidity: one cardiac arrest, one ‘shock-disturbance of consciousness’, and one myocardial infarction in association with the drug. The comments of the manufacturer were: “it seems too early to attribute the reports to gemeprost, because effects of infection, anaesthesia, amniotic fluid embolism, etc. could be the causes” (Nourypharma, personal communication, 1998).

In the two cases presented here, myocardial ischaemia resulted in circulatory collapse. We cannot be absolutely certain of a cause-and-effect relationship, but it seems prudent to assume that—as prostaglandin E2—the prostaglandin E1 (gemeprost) is capable of causing coronary artery spasm in susceptible patients. Whether gemeprost affected cerebral artery tone, as perhaps in the second case as well as coronary, cannot be answered. Cerebral vasoconstriction reducing cerebral blood flow has been suggested as a cause of patients experiencing major convulsions in the course of prostaglandin induced abortion7. However, the resulting cerebral damage observed in the patient presented here could be explained by interruption of cerebral perfusion pressure for some time before the circulatory collapse was detected.

No published data concerning the total usage of gemeprost are available. Therefore, it is not possible to estimate the frequency of myocardial ischaemia caused by the abortifacient. While the providing company Nourypharma stated (personal communication, 1999), that in Germany today about 8500 packages (each including 10 vaginal suppositories) are sold annually, this information cannot be translated to the number of patients actually exposed to the risk of developing coronary artery spasm. In addition, it is not known how many serious myocardial ischaemic events have occurred since 1989, the year gemeprost began to be marketed in Germany. There is potential for an unknown degree of under-reporting.

The observations based on the electrocardiogram, and coronary artery angiography presented in this report provide support for the suggestion that coronary artery spasm was responsible for myocardial infarction in one woman having coronary artery disease and in circulatory collapse followed by cerebral infarction in another, who was—except for a smoking history—devoid of any risk factors for cardio- and cerebrovascular disease. Similar serious cardiovascular incidents led the French health authorities8 to state that smoking and any suspicion of cardiovascular disease are a contraindication to using any type of prostaglandin as an abortifacient. These contraindications should be included as part of the product information, which has not been the case with the German product information.

In response to the serious cases of myocardial ischaemia reported here the German provider of gemeprost in 1999 have revised the warning section of the drug label to warn practitioners of the risk of coronary spasm followed by myocardial infarction.

It seems prudent to monitor blood pressure, pulse rate and the electrocardiogram in every woman during the course of abortion in an environment where corrective treatment may be instituted immediately in the event of cardiovascular complications. This would lead to a better understanding of the cardiovascular effects of prostaglandin E1 when used for abortions or evacuation for fetal death.


The author gratefully acknowledges the assistance of Dr H. Rosenberg of Jefferson Medical College, Philadelphia, Pennsylvania, USA in editing the English text of this case report.